CONCRETE LIBRARY OF JSCE NO. 38, DECEMBER 2001
EFFECTS OF INITIAL HYDRATION REACTIVITY OF CEMENT ON THE DISPERSING
PERFORMANCE OF POLYCARBOXYLATE SUPERPLASTICIZER
-
(Translation from Concrete Research and Technology, Vol. 11, No. 2, May 2000)
Kazuo Yamada
Shunsuke Hanehara
Kenichi Honma
The effects of cement’s initial hydration reactivity on the dispersing performance of a polycarboxylate
superplasticizer (PC) are investigated using normal Portland cement (NPC) and belite-rich low-heat Portland
cement (LHC). Initial hydration reactivities are controlled for both NPC and LHC by exposing them to a
condition of 80%RH and 20°C. Regardless of differences in cement type and initial hydration reactivity of
cement, the fluidity of paste containing PC is found to be positively correlated to the PC adsorption per BET
specific surface area of the solid phase of the paste, Adpg/BEI. This means that the dispersing force of the PC is
determined by the amount of PC adsorbed onto a unit surface area of the solid phase. In both NPC and LHC
cases, AdpC/BET corresponds to the variation in BET specific surface area of the solid phase of the paste, SSAPC.
The variation in SSAPC is caused by the variation in cement initial hydration reactivity. "Despite SSAPC,
experimental results indicate that ACIPC/BET is also affected by the sulfate ion concentration in the aqueous phase
of the paste, because sulfate ions act as competitive adsorbates. The value of Adpc/BET for LHC is larger than
that of NPC at the same SSAPC level because of the lower alkaline sulfate content in LHC.
Key Words: Polyearboxylate superplasticizer; hydration reactivity, fluidity, BET specific surface area, sulfate
ion
'
in
11$
_*
ii
__
R’
1
1&2
‘i
i
‘I
'
I’
R
"
1
Kazuo Yamada is a research scientist in the Cement Chemistry Group of the Central Research & Development
Center of Taiheiyo Cement Corporation, Sakura, Japan. He obtained his Doctor of Engineering degree from
Tokyo Institute of Technology in 2000. His research activities relate to the chemistry of cement, y cement
dispersants, expansive additives for concrete, and the durability of concrete. He is a member of the JSCE and
JCI.
A
Shunsuke Hanehara is a manager in the Technology Management Group of the Central Research &
Development Center of Taiheiyo Cement Corporation, Sakura, Japan. He obtained his Doctor of Engineering
degree from the Keio University in 1986. His research activities relate to the chemistry of cement, the
microstructure of concrete, and the dHI2Pi1iILOf concrete. He is a member of the JSCE and JCI.
Kenichi Honma is a research assistant in the Cement Technology Group of the Central Research &
c
1
‘R
1
“Q
1
1
I
Development Center of Taiheiyo Cement Corporation, Sakura, Japan. His research activities relate to the
__c11em;§try ofpce_r_1i_ent_a_1_1dJtl_1e manufacturing of cement. He is a member of the JCI.
1. INTRODUCTION
described
[2],
[3],
[4],
[5],
[6].
In
particular,
various
experiments have been carried out on the fluidity of cement
pastes made with cement of various characteristics and with
superplasticizers
of different types: polycarboxylate
(PC),
naphthalene sulfonate (NS), and melamine sulfonate (MS).
Based on these experimental results, two parameters have
been proposed for the description of superplasticizer
performance: critical dosage and dispersing ability, as shown
inFig.
1 [7].
The effects of adding a superplasticizer
on the fluidity of a
cement paste can be classified as follows [7]. Up to a certain
dosage, the fluidity does not increase. Once this critical
dosage is reached, fluidity increases in proportion to the
dosage. In this effective region of superplasticizer
addition,
the rise in fluidity per increment in superplasticizer
dosage is
defined as the dispersing ability. At a certain superplasticizer
dosage, fluidity tends to saturate. In the most common
form of
superplasticizer
usage in concrete, the superplasticizer
is used
within the range of dosage in which the fluidity can be
adjusted by the dosage. In Fig. 1, this range corresponds to the
linear part of the relationship.
It is the mechanism of
superplasticizer
operation in this range that will be discussed
in this study.
i
The fluidity of concrete containing a superplasticizer
is
knownto vary unexpectedly with various factors including
the type of cement, differences among manufacturing plant,
factories and others [1], In various attempts to control these
variations in fluidity, the mechanism of interaction between
cement and superplasticizer
has been studied and the
relationship between the characteristics
of cement and the
dispersing
performance of the superplasticizer
has been
53
Fig. 1 Critical
I
I
°-6
I
I
I
0.4
0.2
5 o.o
0
Fig. 2 Initial
critical
-72
-
5
Hydration
10
15
20
heat / J/g, 0-30 min
hydration
reactivity
dosage [7]
and
^ 30°
I 250
X 200
à"<
Q
X. 150
100
I50
0.1
0.2
0.3
R+/ mass%
Fig.
3
Soluble alkaline content and
dispersing ability [7]
12
Increase in [S04
from 0 to 400
"]
o
mmolA
à"J3
æf Na2SO4 added in mixing water
O Na2SO4 added in mixed paste
u
The mechanism by which alkaline
sulfate ions affect
dispersing
performance has been studied by several
researchers [8], [9], [10]. It is assumed that PC is adsorbed
onto the surface of the cement and that this adsorption
proceeds until equilibrium with the coexisting sulfate ions
[11]. In the case of a high sulfate ion concentration, the
adsorption equilibrium moves to the desorbed side, as shown
in Fig. 4, and the dispersing performance of PC decreases.
Similar behavior is exhibited by NS [7]. Here, NS adsorbed
onto the cement surface can be removed by increasing the
sulfate ion concentration, as shown in Fig. 5. This effect can
be demonstrated by an experiment using limestone powder
whose surface is expected to be stable in the adsorption
experiments, as well as the case of the delayed addition of NS
to cement paste. Someresearchers have assumed that the loss
[7]
0.8
à"I
In a correlation analysis of cement characteristics, the critical
dosage indicates a positive relationship with initial hydration
reactivity of the cement for each type of superplasticizer,
as
shown in Fig. 2. On the other hand, the dispersing ability
shows a negative relationship with the alkaline sulfate content
of the cement, as shown in Fig. 3 for each type of
superpla stic izer.
y Saturation dosag
SP dosage
dosage and dispersing
ability
0
0.2
0.4
PC adsorption ratio
Fig. 4 Relationship between PC
adsorption and fluidity [1 1]
0.6
of dispersing performance of NS when the water-cement ratio
is low can be attributed to a reduction in the electric double
layer, which is set up by the adsorption of NS, because ionic
strength rises at low-water cement ratios [12], [13]. However,
a recent study suggests an effect of sulfate ions on the
competitive adsorption equilibrium of NS with sulfate ions
[14].
In the case of NS, one important effect contributes to the
adsorption equilibrium;
the absorption into initial hydrates.
The authors have investigated
the mechanism of this
absorption from experiments using cements with various
hydration reactivities
[15] with reference to previous studies
[2],
[3],
[4],
[5],
[16],
[17],
[18],
0
0. 1
0.2
SO42" concentration
0.3
0.4
(mol/1)
[19].
In the case of PC, there have been few studies of the effect of
initial hydration on dispersing performance. This study is an
investigation
of the mechanisms at work using cements with
various initial hydration reactivities controlled by exposure to
moist conditions in a humid atmosphere.
Table 1 Cement characteristics
Fig. 5 The effect of sulfate ion concentration
on the dispersing performance and
adsorption behavior of NS [7]
(mass%) [20]
*) R2SO4 represents the soluble alkaline sulfate based on measurementof Na+ and K+in a solution exracted by
centrifugal separation a cement paste with a water-cement ratio of 40 % immediately after mixing.
2. Samples and experimental method
2.1 Samples
The cements used in the experiment were commercial normal Portland cement (NPC) and high-belite low-heat
cement (LHC) as specified by JIS R 5201 and JIS R 5205, respectively. The excellent compatibility
of LHC
with PC means that it is commonly in concrete with a low water-cementratio. The reason for this excellent
compatibility
is thought to be the low alkaline sulfate content as well as the low C3Acontent of LHC [21].
The initial hydration reactivity of cement samples was modified by exposing them to moist conditions. To do
this, a vessel measuring 7 cm in depth, 25 cm in width, and 35 cm in length was filled with 5 kg of cement and
exposed to an atmosphere characterized by 20 °C and 80 %RH for 0.25, 1, 3, and 7days. The basic
characteristics
of these cements have been previously reported [20]. Three-day exposure to these conditions
corresponds to storage of bagged cement in a warehouses for 8 weeks.
The superplasticizer
used was a commercial polycarboxylate
superplasticizer
with polyether graft chains. Its
solid content was 27.1 mass% after drying at 105°C. A chemical analysis was carried out to gain an
understanding of the chemical structure of the PC. The graft chain consisted of polyoxyethylene
comprising 23
repetitions of ethyleneoxide. The trunk chain was polymerized methacrylate including matharylsulfonate.
The
molecular weight was 47,000 (mass average) as measured with a gel permeation chromatograph using a column,
Shodex OHpak SB806M. Polyoxyethylene
was used as the standard material. The molar ratio of each
component was 66 mol% of methacrylic acid, 21 mol% of grafted methacrylate, and 6 mol% of
metharysulfonate. For reference, a beta-naphthalene sulfonate superplasticizer
(NS) was also used.
-73
-
2.2 Properties of cement paste
The dispersing performance of the superplasticizer
was evaluated by measuring the fluidity of a cement paste.
The superplasticizers
were added to the cement in the form of a premixture with water. Cement paste with a
30% water-cementratio was mixed in a Hobart mixer for 1 min. at low speed followed by 3 mins. at high speed.
The flow (f) of the cement paste was measuredjust after mixing in accordance with JASS 15 M 103-3.5 using a
pipe 50 mmof diameter and 51 mmof height. As an index of cement paste fluidity, the relative flow area was
calculated as follows:
f
p=
-502
502
2.3 Analysis of aqueous paste phase
The aqueous phase was separated in a centrifugal separator (rotating rate: 13,5000 rpm; rotating radius: 12 cm;
and centrifuge time: twice for 1 min.) just after mixing. The sulfate ion concentration in the aqueous phase of
the paste was measured using an ion chromatograph. The PC concentration was measured using a total organic
carbon analyzer. By comparing the PC concentrations before and after mixing, the change in PC concentrations
wasobtained and assumed to correspond to the amount adsorbed.
2.4 Analysis of initial
hydration
reactivity
The initial hydration reactivity of the cement was evaluated by measuring the heat of hydration in a micro
calorimeter for 30 min. after mixing with water. The specific surface area of the solid phase of the hydrated
cement paste was also measured. First, after mixing the cement paste, it was immersed in plenty of acetone in
order to stop hydration. The solid phase was then separated by suction filtration. The separated sample was
dried at 20°C and 11%RH for 7 days and the specific surface area was then measured by the one^-poirit BET
method using nitrogen gas adsorption.
Table 2 Changes in cement characteristics
after exposure to moist conditions
[20]
3. Results and discussions
3. 1 Change in cement characteristics
after exposure to moist conditions
The changes in cement characteristics
after exposure to moist conditions are summarized in Table 2 [20],
Although the Blaine specific surface area did not change, the BET specific surface area increased as the period
of exposure increased, rising by 1.4 times from that of untreated cement after 7 days of exposure. Observation
with a SEM (scanning electron microscope) revealed that fine hydrate particles measuring less than 1 nm were
formed as a result of exposure to moist conditions. These fine particles were identified as probably ettringite
and gypsum in XRD (X-ray diffraction analysis).
As for the initial hydration reactivity, the heat released in the 30 min. after mixing with water decreased from
14.1 J/g for unexposed NPC to 2.0 J/g after 7 days of exposure. It decreased from ll.5 J/g for unexposed LHC
to 1.8 J/g after the same period of exposure. The heat release in this period is thought to be attributable to the
hydration reaction of interstitial
phases such as C3A to ettringite. Comparing the release of hydration heat by
LHC and NPC, the amount of C3A in LHC is half that in NPC, as shown in Table 2, whereas the amount of
hydration heat only 1 8 % less. This result indicates that the release of hydration heat is not directly related to the
74 -
25
20
20
V
o
25
15
15
10
10
-æfNPC
5
5
-O»
LHC
0
0
0
0
2
4
Exposure time (day)
Fig. 6 Effect of exposure on fluidity
2
6
Fig. 7 Effects
(NS)
4
6
Exposure time (day)
of exposure on fluidity
(PC)
amount of C3A. Thus it is necessary to evaluate the amount of C3A that takes parts in reaction.
Exposure to moist conditions is shown to successfully
be discussed in the next section.
control the initial
hydration
reactivity,
and this effect will
3.2 Change in fluidity of paste by exposure of cement to moist conditions
The changes in fluidity of cement paste containing superplasticizers
due to exposure to moist conditions are
shown in Figs. 6 and 7 for NS and PC, respectively. The superplasticizer
dosages were 1.2 and 0.8 mass% per
kg of cement as the received forms of commercial liquid for NS and PC, respectively.
When unexposed, NS indicated a fluidity of T= 13 for NPC and T= 3 for LHC. The fluidity of paste with NS
increased for both NPC and LHC to the same value of T= 18 after exposure to moist conditions; this took one
day in the case ofNPC and more than three days in the case ofLHC.
Further, the fluidity of NPC paste with PC showed a similar tendency to the NS case, with the fluidity
increasing after exposure to moist conditions up to F= 15~18 after more than 0.25 days. In contrast, the
fluidity of LHC paste showed a different tendency. Although the fluidity was high in the unexposed state (F=
24), it decreased to F= 14 after one day of exposure to moist conditions. The fluidity then remained constant
with longer exposure to moist conditions.
For the case of LHC, which showed significant change in fluidity with exposure to moist conditions and
behavior dependent on the type of superplasticizer,
the relationship between dosage and fluidity was examined.
Relationships
for unexposed and 7-day exposed LHC in the case of mixing with NS and PC are shown in Figs.
8 and 9, respectively.
In the case of NS, the critical dosage decreased by half and dispersing ability increased with exposure to moist
conditions. As a result, for a constant dosage of NS, the fluidity of paste with NS can be expected to increase
with exposure to moist conditions. In the case of PC, the critical dosage changed little with exposure to moist
conditions because the critical dosage was not very significant even in the unexposed case. Dispersing ability
decreased with exposure to moist conditions. As a result, in the case of constant dosage of PC, the fluidity of
PC-dosed paste can be expected to decrease with exposure to moist conditions.
3.3 Mechanism of change in fluidity with NS
The mechanism by which the fluidity of paste including NS changes with variations in initial hydration
reactivity of the cement is discussed here with reference to a previous paper [20]. Besides adsorption of NS onto
cement hydrates, a significant amount of NS is thought to be absorbed into the hydrates in competition with
sulfate ions. This absorbed NS does not exhibit a dispersing effect. As shown in Table 2, as the initial hydration
reactivity of cement is reduced through exposure to moist conditions, the amount of hydrates produced
decreases and NS tends to remain in the aqueous phase. There, it contributes further to dispersion. It is thought
to be for this reason that the critical dosage of NS decreases.
75
25
25
20
20
15
15
10
10
5
5
0
0
0.5
1
0.5
1.5
NS dosage (mass%)
Fig. 8 Effect of exposure on relationship between NS
dosage
,
--0--LHC-no-SP
1?6
*
and fluidity
J
-ALHC-NS
-(a)
-7
1.5
PC dosage (mass%)
*& 9 mect of exPosure on relationshiP between NS
, a....
and fluidity
dosage
1
-ANPC-NS
-æfNPC-no-SP
k
55
2<f
*°--o
0
o
i
?
2
4
2
6
4
6
Exposure (day)
Exposure (day)
Fig. 10 Effect of exposure on specific surface area (S.S.A.) of initial
2
4
6
Exposure (day)
hydrates
The change in specific surface area of the solid phase of the paste is shown in Fig. 10. The BET specific surface
area of the solid phase of paste dosed with NS (abbreviated as SSANS) increased in the unexposed case as
compared to that of paste without the superplasticizer.
With exposure to moist conditions, SSANs decreased
irrespective of the kind of cement. In general, a superplasticizer
is thought to exhibit a dispersing effect only
after adsorption onto particles, and the dispersing performance is proportional to the amount adsorbed per unit
surface area. Therefore, for a particular NS dosage, dispersing ability increases because the surface on which
the NS is adsorbed decreases; that is, the NS adsorption per unit surface area of the hydrates increases. For a
moredetailed discussion of this issure, it is necessary to make a distinction between the NS absorbed by
hydrates and the NS adsorbed onto hydrates.
3.4 Mechanism of change in fluidity
with PC
a) Subject of analysis
The initial hydration reactivity of cement did not affect the critical dosage very much in the case of PC.
Consequently, the dispersing ability is discussed here with reference to Fig. 9. Where the critical dosage is
constant and independent of exposure to moist conditions, a comparison of T at constant PC dosage corresponds
to a comparison of dispersing ability. Thus the analysis described in this section focuses on the T change of
cement paste with a constant PC dosage, as shown in Fig. 7, in order to investigate the mechanism by which
fluidity changes with exposure to moist conditions.
b) PC adsorption and fluidity
The change in PC adsorption per mass of the solid phase paste mixed with PC when exposed to moist
conditions is shown in Fig. 1 1. LHC adsorbed more PC than NPC. Adsorption decreased with exposure up to
three days and thereafter increased slightly.
~76 -
The relationship
between the amount of PC adsorbed and T is shown in
Fig.12. In the case of LHC, there was a positive relationship between them.
However, in the case of NPC, the relationship
was negative. Thus the
behavior differed with the type of cement, and the change in T cannot be
explained simply in terms of PC adsorption.
c) Effect of surface area of solid phase of paste
The dispersing performance of PC is thought to be more significantly
affected by the adsorption per unit surface area of the solid phase
(Adpc/BEi) than by the actual amount of PC adsorbed. To investigate this,
the change in BET specific surface area of the solid phase of paste
containing PC (SSApc) with exposure to moist conditions was examined.
The results are shown in Fig. 10(c). When compared with the unexposed
cement paste without superplasticizer
for each cement type, SSApc was
higher in the case of NPC and was almost the same in the case ofLHC. In
the NPC case, SSApc decreased with exposure time, as in the case of NS.
In the LHC case, SSAPC reached a maximumvalue at 1 day and decreased
with further exposure to moist conditions.
It is clear that changes in BET specific surface area after exposure to moist
conditions
varies significantly
for different
combinations
of
superplasticizer
and cement. Although detailed analysis of the reasons for
of such variations is outside the scope of this study, the authors speculate
that the addition of a superplasticizer
may affect the initial hydration of the
cement, and this effect may differ for each type of superplasticizer.
Within
the time scale of this study, the hydration reactions that need to be
considered are the dissolution
of calcium sulfate hemihydrate, the
deposition
of gypsum, the hydration of free lime, the formation of
calcium-sulfo-aluminate
hydrates such as ettringite by the hydration of
C3A, and similar reactions. In a system without a superplasticizer,
it has
been reported that the phase composition of hydrates is altered by
exposure to moist conditions
[20]. Also, Kang et al. [21] have
demonstrated that PC significantly
affects the hydration of C3A. Changes
in initial
hydration reactivity
and the presence or otherwise of a
superplasticizer
can both affect the production of initial hydrates, and the
effects may be different for NPC and LHC. In this study, the complex
changes in the surface area of hydrates resulting from a combination of
these two effects were observed. As for the mechanism of this
phenomenon, further study will be necessary to elucidate moredetails.
0
2
Next, the controlling factor of Adpc/BETis investigated.
Because the PC
dosage was constant, SSApc is examined first. The relationship between
SSApc and Adpc/BETis shown in Fig. 14. As SSApc decreased, Adpc/BET
tended to increase with both NPC and LHC.Again, because the PC dosage
was constant, the PC adsorbed per unit surface area increased as the
surface area of solid phase on which the PC is adsorbed fell.
d) Effect of sulfate ion concentration in aqueous phase
Figure 14 compares Adpc/BETfor NPC and LHC at the same SSApc. The
value of Adpc/BETfor NPC is lower than that for LHC. The mechanism
-77
-
6
25
20
15
10
5
0
0
0.5
1
AdpC (mg/g)
Fig. 12 Relationshp between PC
adsorption and fluidity
25
20
15
10
5
0
0.0
0.2
0.4
Adpc/BET (mg/m2)
Fig.
The relationship between Adpc/BETand F is shown in Fig. 13. The same
linear relationship
between Adpc/BETand F is seen irrespective of the type
of cement and length of exposure to moist conditions. This suggests that
the fluidity of a paste is determined by AdPC/BET,which in turn means that
the dispersing
performance of a superplasticizer
is determined by the
amount of superplasticizer
per surface area of the particles in the
dispersing phase, as noted above.
4
Exposure (day)
Fig.1 1 Effect of exposure on
adsorption of PC
13 Adpc/BET and fluidity
0.6
behind this phenomenon is examined in terms of adsorption
equilibrium.
S" 140
o
The PC is thought to be in adsorption equilibrium at the
surface of the cement hydrates in competition with sulfate
ions. This implies that the adsorption state is affected by
sulfate ion concentration. To check this, the sulfate ion
concentration was measured, and the results are shown in Fig.
15. The aqueous phase of a cement paste is saturated state
with lime and gypsum, and the sulfate ion concentration is
strongly affected by the alkaline content, as determined by the
content of alkaline sulfates in the cement. Because the
alkaline sulfate content of LHC is lower than that of NPC, as
shown in Table 1, the sulfate ion concentration is lower in the
case of LHC than in the case of NPC. As for the effect of
exposure to moist conditions,
sulfate ion concentration
increased for first 0.25 days of exposure in the case of both
cements. With lengthier exposure, the change in sulfate ion
concentration was limited. The initial difference in sulfate ion
concentrations of LHC and NPC is three times greater than
the change seen as a result of exposure to moist conditions.
I
|
Based on these results, the reason for Adpc/BETbeing smaller
in the case of NPC than in the case of LHC can be discussed.
As mentioned above, the PC dosage was constant, so Adpc/BET
increased when the surface area was small. The size of this
increase depends on the sulfate ion concentration, because PC
competes with sulfate ions for adsorption. In the case of a
higher sulfate ion concentration, the amount of PC adsorbed
on the solid surface is less, even where the PC dosage is
constant. As a consequence of this mechanism, the Adpc/BEr
value of NPC, which has a higher sulfate ion concentration, is
less than that of LHC, which that has a lower sulfate ion
concentration.
The effect of sulfate ion concentration in the aqueous phase of
the paste is shown in Fig. 16. This figure, based on published
data [23], shows the change in PC adsorption in a PC-dosed
NPC paste (PC dosage = 0.8 mass%; water-cement ratio =
0.30) when the sulfate ion concentration is altered by the
addition of soluble calcium salt to the mixed cement paste. By
adding calcium salt in this manner, it can be assumed that the
surface area of the solid phase of the paste remains constant,
so the data for PC adsorption in reference
[23] were
converted to Adpc/BETby dividing by the SSApc values of
unexposed NPC paste.
|
120
10°
80
I60
8
40
J£ . 0
03
0
2
4
6
8
Exposure (day)
Fig. 1 5 Change in sulfate ion concentration
in aqueous phase of paste with PC
with exposure
0
50
100
150
Sulfate ion cone. (mmol/1)
Fig. 16 Effect of sulfate ion concentration
PC adsorption per BET S.S.A.
on
0.4
0.3
0.2
0.1
0.0
0
2
S SAPC(m2/g)
4
Fig. 17 Relationshp
between SSAPC and
Adpc/BEx with consideration
of
difference in sulfate ion concentration
Based on this data, the Adpc/BET values in Fig. 14 were
normalized to a constant sulfate ion of 100 mmol/1. That is the
ratio of AcIpc/bet at two different sulfate ion concentrations
wasused for the normalization;
for example, Adpc/BETwas 0.31 mg/m2 when the sulfate ion concentration was
50 mmol/1 and 0.25 mg/m2 when the sulfate ion concentration was 100 mmol/1, so the ratio is 1.25. All data in
Fig. 14 were normalized to the case of 100 mmol/1 of sulfate ion concentration. The results are shown in Fig.
17.
In this figure, the difference between NPC and LHC is less than in Fig. 14. This means that the reason for the
lower Adpc/BETvalue of NPC compared with LHC at the same SSApc can be attributed to the difference in
sulfate ion concentration. However, the behavior of NPC and LHC does not coincide exactly even in Fig. 17.
-78
This may be explained by ambiguity in the effect of sulfate ions on PC adsorption behavior when the surface
area and the phase composition of the hydrates are modified with exposure to moist conditions. In other words,
it can be expected that the hydrate composition would change through exposure of the cement to moist
conditions, and the effect of sulfate ions on PC adsorption onto each different hydrate may be different.
e) Mechanism by which initial hydration reactivity affects dispersing performance of PC
Based on the discussion above, the mechanism by which the initial hydration reactivity of the cement affects PC
performance is summarized. The fluidity of paste dosed with PC is determined by Adpc/BETregardless of the
cement type and initial hydration reactivity. Since PC acts as a dispersant immediately upon adsorption onto the
solid phase, SSApc plays an important role. Changes in the initial hydration reactivity cause variations in SSApc,
and these in turn cause a change in Adpc/BETunder conditions of constant PC dosage. On the other hand, PC
adsorption on the solid phase is in a state of equilibrium determined by the balance of PC and sulfate ion
concentrations in the aqueous phase. Consequently, differences in the relationships
between SSApc and
Adpc/BETin the cases of NPC and LHC are thought to derive from difference in sulfate ion concentration. In
short, in order to discuss the mechanism by which initial hydration reactivity affects the performance of PC, it is
important to consider changes in the specific surface area of the solid phase of the cement paste and in the
concentration of sulfate ions in the aqueous phase of the paste, and these changes result from variations in the
initial hydration reactivity of the cement.
4. Conclusion
Initial hydration reactivity affects the dispersing performance of PC. The mechanism of this behavior was
investigated
from the viewpoint of PC adsorption behavior. The cements used in the investigation were NPC
and LHC, and initial hydration reactivity was controlled with exposure of the cement to moist conditions. NS
wasalso examined by referring to the literature.
1)
2)
3)
4)
In the case of paste containing NS, fluidity increased with exposure to moist conditions irrespective of the
type of cement. This occurs because the critical dosage falls and the dispersing ability rises. In the case of
PC, although the fluidity of NPC paste increased slightly after exposure to moist conditions, the fluidity of
LHC decreased significantly.
This decrease in the LHC case derives from a fall in dispersing ability and the
fact that the critical dosage remains constant even whenthe initial hydration reactivity falls.
The increased fluidity of paste posed with NS after exposure to moist conditions is thought to be caused by
the higher Adpc/BET,which in turn results from reduced absorption into initial hydrates and the lower SSAns
caused by lower initial hydration reactivity.
The fluidity of paste dosed with PC exhibits a single positive correlation with Adpc/BET irrespective of
cement type and initial hydration reactivity. This result demonstrates that the effect of initial hydration
reactivity on Adpc/BETis important in analyzing the influence of initial hydration reactivity on PC dispersing
performance.
Although Adpc/BETwas affected by changes in SSApc caused by changes in the initial hydration reactivity
of cement, NPC showed lower Adpc/BETthan LHC for the same SSApc. This difference can be explained to
somedegree by considering differences in sulfate ion concentration. Because sulfate ions and PC compete
for adsorption onto the surface of the initial hydrates, and because of the higher sulfate ion concentration in
the aqueous phase of the NPC paste than in LHC-paste, AdpcmETof NPC is lower than that of LHC.
References
[1] Hanehara, S. and Yamada, K., "Interaction between cement and chemical admixture from the point of
cement hydration, admixture adsorption and paste rheology", Cement and Concrete Research, Vol. 29, No.
8, pp.
1159-1165,
1999
[2]
Hattori, K., Okada, E. and Mizumuna, T, "Characteristics
of normal Portland cement and its dispersion
by
superplasticizer",
JCA Proceedings of Cement & Concrete, No. 38, pp. 122-125, 1984 (in Japanese)
[3] Aictin,
P. C, Jolicoeur, C. and MacGregor, J. G., "Superplasticizers:
How they work and why they
occasionally
don't", Concrete International,
Vol. 5, pp. 45-52, 1 994
[4] Yamaguchi, O., Nakajima, Y, Sugaya, H. and Takahashi, M., "The fluidity of pastes and mortars containing
various types of water-reducing
agent", JCA proceedings of Cement & Concrete, No. 49, pp. 216-221,
1 995 (in Japanese)
[5] Nawa, T, Eguchi, H. and Hukaya, Y, "Effect of alkali sulfate on the rheological
behavior of cement paste
-79
-
containing
International
a superplasticizer",
in: Malhotra, V. M. (ed.), "Proceedings of the 3rd CANMET/ACI
Conference on Superplasticizers
and Other Chemical Admixtures in Concrete", ACI, SP-1 19,
pp. 405-424,
1989
[6] Yamada, K., Hanehara, S. and Matsuhisa, M., "Twoworking parameters representing the dispersing
mechanism of a superplasticizer
-Critical dosage and dispersing
ability-",
Concrete Research and
Technology, Vol. 10, No. 3, pp. 61-68, 1999 (in Japanese)
[7] Yamada, K., Ogawa, S. and Hanehara, S., "Working mechanism of poly-beta-naphthalene
sulfonate and
polycarboxylate
superplasticizer
types from the point of cement paste characteristics",
V. M. Malhotra (ed.),
6th CANMET/ACI International
Conference on Superplasticizers
and Other Chemical Admixtures in
Concrete,
SP1 95, ACI, pp.367-382,
2000
[8]
Kato, H. and Yoshioka, K., "Effect of sulfate ions on the adsorption behavior and dispersing performance of
superplasticizers",
JCA Proceedings of Cement & Concrete, No. 53, pp. 115-1 17, 1999 (in Japanese)
[9] Coppola, L., Troli, R., Zaffaroni, P. and Collepardi, M., "Influence of the sulfate level in the clinker phase
on the performance of superplasticized
concretes", in: Malhotra, V. M. (ed.), "4th CANMET/ACI/JI Inter.
Symp. on Advances in Concrete Technology" (ACI, SP-179), pp. 271-281, 1998
[10]Negishi,
K., Nakajima, Y, Sugaya, H. and Goto, T., "Effects of polymorphism of calcium sulfate and
soluble alkalis on the fluidity of cement paste containing polycarboxylic
acid superplasticizer",
Cement
Science and Concrete Technology, No. 52, pp. 152 - 175, 1998 (in Japanese)
[ll]Yamada,
K., Hanehara, S. and Matsuhisa, M, "Fluidizing
mechanism of cement paste including
polycarboxylate-type
superplasticizer
analyzed from the view point of adsorption
behavior of
superplasticizer,
Proceedings of JCI, Vol. 20, No. 2, pp. 73-78, 1998 (in Japanese)
[12]Satoh,
H, Kitagawa, K and Yamato, E, "The chemical structure and performances of a new type of high
range AE water-reducing agent", JCA Proceedings of Cement & Concrete, No. 49, pp. 222-227, 1995 (in
Japanese)
[13]Nawa, T, "Effects of cement characteristics
on the fluidity of mortar containing a dispersant", Doctor ate
Thesis, Tokyo Institute of Technology, 1992 (in Japanese)
[14]Yamada, K., "Study on working mechanism of polyether-type
cement dispersant based on early cement
hydration and composition variation of solution phase", Doctorate Thesis, Tokyo Institute of Technology,
2000 (in Japanese)
[1 5]Yamada, K., Hanehara, S. and Honma,K., "The effects of naphthalene sulfonate type and polycarboxylate
type superplasticizers
on the fluidity of belite-rich
cement concrete", K. Ozawa and M. Ouchi (ed.),
Proceedings of the International
Workshop on Self-Compacting Concrete, Concrete Engineering Series,
JSCE,
Vol.
30, pp.
201-210,
1988.
[1 6]Massazza, E and Costa, U., "Effect of superplasticizer
on C3A hydration", Proceedings of 7th International
Congress on the Chemistry of Cement, Paris, IV, pp. 529 - 534, 1980
[17]Bonen,
D. and Sarkar, S. L., "The superplasticizer
adsorption capacity of cement pastes, pore solution
composition, and parameters affecting flow loss", Cement and Concrete Research, Vol. 25, No. 7, pp.
1423-1434,
1995
[1 8]Uchikawa,
H, Sawaki, D. and Hanehara, S. "Influence of kind and added timing of organic admixture on
the composition, structure and property of fresh cement paste", Cement and Concrete Research, Vol. 25, No.
2, pp. 353-364,
1995
[19]Nakajima,
Y, Negishi, K. and Goto, T, "Influence of calcium sulfate on early hydration of cement paste
containing superplasticizer",
Cement Science and Concrete Technology, No. 51, pp. 288-293, 1997 (in
Japanese)
[20]Yamada, K., Hanehara, S. and Honma, K., "Change in hydraulic reactivity of cement by aeration and false
setting", Concrete Research and Technology, Vol. 9, No. 2, pp. 91-99, 1998 (in Japanese)
[21]Yamada,
K., Ozu, H. and Yano, M., "Adsorption behavior of polycarboxylate-type
superplasticizer
on
cement hydrates and the effect of sulfate ions", Cement Science and Concrete Technology, No. 54, pp.79-86,
2000 (in Japanese)
[22]Kang, J., Ohba, Y, Sakai, E. and Daimon, M., "Influence of superplasticizers
on hydration in the Ca3Al2Og
CaSO4.2H2O - CaOsystem", Cement Science and Concrete Technology, No. 54, pp.22-27, 2000 (in
Japanese)
[23]Yamada,
K., Ogawa, S. and Hanehara, S., "Controlling
the adsorption and dispersing
force of
polycarboxylate-type
superplasticizer
by sulfate ion concentration in aqueous phase", Cement and Concrete
Research,
Vol. 31, No. 3, pp. 375-384,
2001
-80
-